1. Field of the Invention
The present invention relates to hearing methods and systems. More specifically, the present invention relates to methods and systems that have improved high frequency response that improves the speech reception threshold (SRT) and preserves and transmits high frequency spatial localization cues to the middle or inner ear. Such systems may be used to enhance the hearing process with normal or impaired hearing.
Previous studies have shown that when the bandwidth of speech is low pass filtered, that speech intelligibility does not improve for bandwidths above about 3 kHz (Fletcher 1995), which is the reason why the telephone system was designed with a bandwidth limit to about 3.5 kHz, and also why hearing aid bandwidths are limited to frequencies below about 5.7 kHz (Killion 2004). It is now evident that there is significant energy in speech above about 5 kHz (Jin et al., J. Audio Eng. Soc., Munich 2002). Furthermore, hearing impaired subjects, with amplified speech, perform better with increased bandwidth in quiet (Vickers et al. 2001) and in noisy situations (Baer et al. 2002). This is especially true in subjects that do not have dead regions in the cochlea at the high frequencies (Moore, “Loudness perception and intensity resolution,” Cochlear Hearing Loss, Chapter 4, pp. 90-115, Whurr Publishers Ltd., London 1998). Thus, subjects with hearing aids having greater bandwidth than the existing 5.7 kHz bandwidths can be expected to have improved performance in quiet and in diffuse-field noisy conditions.
Numerous studies, both in humans (Shaw 1974) and in cats (Musicant et al. 1990) have shown that sound pressure at the ear canal entrance varies with the location of the sound source for frequencies above 5 kHz. This spatial filtering is due to the diffraction of the incoming sound wave by the pinna. It is well established that these diffraction cues help in the perception of spatial localization (Best et al., “The influence of high frequencies on speech localization,” Abstract 981 (Feb. 24, 2003), Association for Research in Otolaryngology). Due to the limited bandwidth of conventional hearing aids, some of the spatial localization cues are removed from the signal that is delivered to the middle and/or inner ear. Thus, it is oftentimes not possible for wearers of conventional hearing aids to accurately externalize talkers, which requires speech energy above 5 kHz.
The eardrum to ear canal entrance pressure ratio has a 10 dB resonance at about 3.5 kHz (Wiener et al. 1966; Shaw 1974). This is independent of the sound source location in the horizontal plane (Burkhard and Sachs 1975). This ratio is a function of the dimensions and consequent relative acoustic impedance of the eardrum and the ear canal. Thus, once the diffracted sound wave propagates past the entrance of the ear canal, there is no further spatial filtering. In other words, for spatial localization, there is no advantage to placing the microphone any more medial than near the entrance of the ear canal. The 10 dB resonance is typically added in most hearing aids after the microphone input because this gain is not spatially dependent.
Evidence is now growing that the perception of the differences in the spatial locations of multiple talkers aid in the segregation of concurrent speech (Freyman et al. 1999; Freyman et al. 2001). Consistent with other studies, Carlile et al., “Spatialisation of talkers and the segregation of concurrent speech,” Abstract 1264 (Feb. 24, 2004), Association for Research in Otolaryngology, showed a speech reception threshold (SRT) of −4 dB under diotic conditions, where speech and masker noise at the two ears are the same, and −20 dB with speech maskers spatially separated by 30 degrees. But when the speech signal was low pass filtered to 5 kHz, the SRT decreased to −15 dB. While previous single channel studies have indicated that information in speech above 5 kHz does not contribute to speech intelligibility, these data indicate that as much as 5 dB unmasking afforded by externalization percept was much reduced when compared to the wide bandwidth presentation over virtual auditory simulations. The 5 dB improvement in SRT is mostly due to central mechanisms. However, at this point, it is not clear how much of the 5 dB improvement can be attained with auditory cues through a single channel (e.g., one ear).
It has recently been described in P. M. Hofman et al., “Relearning sound localization with new ears,” Nature Neuroscience, vol. 1, no. 5 Sep. 1998, that sound localization relies on the neural processing of implicit acoustic cues. Hofman et al. found that accurate localization on the basis of spectral cues poses constraints on the sound spectrum, and that a sound needs to be broad-band in order to yield sufficient spectral shape information. However, with conventional hearing systems, because the ear canal is often completely blocked and because conventional hearing systems often have a low bandwidth filter, such conventional systems will not allow the user to receive the three-dimensional localization spatial cues.
Furthermore, Wightman and Kistler (1997) found that listeners do not localize virtual sources of sound when sound is presented to only one ear. This suggests that high-frequency spectral cues presented to one ear through a hearing device may not be beneficial. Martin et al. (2004) recently showed that when the signal to one ear is low-pass filtered (2.5 kHz), thus preserving binaural information regarding sound-source lateral angle, monaural spectral cues to the opposite ear could correctly interpret elevation and front-back hemi-field cues. This says that a subject with one wide-band hearing aid can localize sounds with that hearing aid, provided that the opposite ear does not have significant low-frequency hearing loss, and thus able to process inter-aural time difference cues. The improvement in unmasking due to externalization observed by Carlile et al. (2004) should at least be possible with monaural amplification. The open question is how much of the 5 dB improvement in SRT can be realized monaurally and with a device that partially blocks the auditory ear canal.
Head related transfer functions (HRTFs) are due to the diffraction of the incoming sound wave by the pinna. Another factor that determines the measured HRTF is the opening of the ear canal itself. It is conceivable that a device in the ear canal that partially blocks it and thus will alter HRTFs, can eliminate directionally dependent pinna cues. Burkhard and Sachs (1975) have shown that when the canal is blocked, spatially dependent vertical localization cues are modified but nevertheless present. Some relearning of the new cues may be required to obtain benefit from the high frequency cues. Hoffman et al. (1998) showed that this learning takes place over a period of less than 45 days.
Presently, most conventional hearing systems fall into at least three categories: acoustic hearing systems, electromagnetic drive hearing systems, and cochlear implants. Acoustic hearing systems rely on acoustic transducers that produce amplified sound waves which, in turn, impart vibrations to the tympanic membrane or eardrum. The telephone earpiece, radio, television and aids for the hearing impaired are all examples of systems that employ acoustic drive mechanisms. The telephone earpiece, for instance, converts signals transmitted on a wire into vibrational energy in a speaker which generates acoustic energy. This acoustic energy propagates in the ear canal and vibrates the tympanic membrane. These vibrations, at varying frequencies and amplitudes, result in the perception of sound. Surgically implanted cochlear implants electrically stimulate the auditory nerve ganglion cells or dendrites in subjects having profound hearing loss.
Hearing systems that deliver audio information to the ear through electromagnetic transducers are well known. These transducers convert electromagnetic fields, modulated to contain audio information, into vibrations which are imparted to the tympanic membrane or parts of the middle ear. The transducer, typically a magnet, is subjected to displacement by electromagnetic fields to impart vibrational motion to the portion to which it is attached, thus producing sound perception by the wearer of such an electromagnetically driven system. This method of sound perception possesses some advantages over acoustic drive systems in terms of quality, efficiency, and most importantly, significant reduction of “feedback,” a problem common to acoustic hearing systems.
Feedback in acoustic hearing systems occurs when a portion of the acoustic output energy returns or “feeds back” to the input transducer (microphone), thus causing self-sustained oscillation. The potential for feedback is generally proportional to the amplification level of the system and, therefore, the output gain of many acoustic drive systems has to be reduced to less than a desirable level to prevent a feedback situation. This problem, which results in output gain inadequate to compensate for hearing losses in particularly severe cases, continues to be a major problem with acoustic type hearing aids. To minimize the feedback to the microphone, many acoustic hearing devices close off, or provide minimal venting, to the ear canal. Although feedback may be reduced, the tradeoff is “occlusion,” a tunnel-like hearing sensation that is problematic to most hearing aid users. Directly driving the eardrum can minimize the feedback because the drive mechanism is mechanical rather than acoustic. Because of the mechanically vibrating eardrum, sound is coupled to the ear canal and wave propagation is supported in the reverse direction. The mechanical to acoustic coupling, however, is not efficient and this inefficiency is exploited in terms of decreased sound in the ear canal resulting in increased system gain.
One system, which non-invasively couples a magnet to tympanic membrane and solves some of the aforementioned problems, is disclosed by Perkins et al. in U.S. Pat. No. 5,259,032, which is hereby incorporated by reference. The Perkins patent discloses a device for producing electromagnetic signals having a transducer assembly which is weakly but sufficiently affixed to the tympanic membrane of the wearer by surface adhesion. U.S. Pat. No. 5,425,104, also incorporated herein by reference, discloses a device for producing electromagnetic signals incorporating a drive means external to the acoustic canal of the individual. However, because magnetic fields decrease in strength as the reciprocal of the square of the distance (1/R2), previous methods for generating audio carrying magnetic fields are highly inefficient and are thus not practical.
While the conventional hearing aids have been relatively successful at improving hearing, the conventional hearing aids have not been able to significantly improve preservation of high-frequency spatial localization cues. For these reasons it would be desirable to provide an improved hearing systems.
2. Description of the Background Art
U.S. Pat. Nos. 5,259,032 and 5,425,104 have been described above. Other patents of interest include: U.S. Pat. Nos. 5,015,225; 5,276,910; 5,456,654; 5,797,834; 6,084,975; 6,137,889; 6,277,148; 6,339,648; 6,354,990; 6,366,863; 6,387,039; 6,432,248; 6,436,028; 6,438,244; 6,473,512; 6,475,134; 6,592,513; 6,603,860; 6,629,922; 6,676,592; and 6,695,943. Other publications of interest include: U.S. Patent Publication Nos. 2002-0183587, 2001-0027342; Journal publications Decraemer et al., “A method for determining three-dimensional vibration in the ear,” Hearing Res., 77:19-37 (1994); Puria et al., “Sound-pressure measurements in the cochlear vestibule of human cadaver ears,” J. Acoust. Soc. Am., 101(5):2754-2770 (May 1997); Moore, “Loudness perception and intensity resolution,” Cochlear Hearing Loss, Chapter 4, pp. 90-115, Whurr Publishers Ltd., London (1998); Puria and Allen “Measurements and model of the cat middle ear: Evidence of tympanic membrane acoustic delay,” J. Acoust. Soc. Am., 104(6):3463-3481 (December 1998); Hoffman et al. (1998); Fay et al., “Cat eardrum response mechanics,” Calladine Festschrift (2002), Ed. S. Pellegrino, The Netherlands, Kluwer Academic Publishers; and Hato et al., “Three-dimensional stapes footplate motion in human temporal bones,” Audiol. Neurootol., 8:140-152 (Jan. 30, 2003). Conference presentation abstracts: Best et al., “The influence of high frequencies on speech localization,” Abstract 981 (Feb. 24, 2003), Association for Research in Otolaryngology, and Carlile et al., “Spatialisation of talkers and the segregation of concurrent speech,” Abstract 1264 (Feb. 24, 2004), Association for Research in Otolaryngology. 